In 1995, 11.3 million women in the United States had urinary tract infection (UTI) with associated costs of approximately $1.5 billion. Escherichia coli accounts for about 90 percent of all UTI present in ambulatory patients. Several different pathogenic processes probably underlie E. coli UTI, analogous to the several different mechanisms that explain diarrheas caused by E. coli. A limited number of E. coli UTI virulence genes have been identified. Known genes in E. coli UTI were identified primarily by phenotypic expression of traits (hemagglutination, toxic activity on tissue culture cells, etc.) hypothesized to be important for colonization or virulence, and then tested by epidemiologic correlations or animal model studies. The goal of this study is to identify additional genes involved in E. coli UTI virulence, transmission, or duration of colonization. The investigators propose the use of a novel approach involving a four stage process, as follows: 1) epidemiologic pairing of bacterial strains with the highest potential to identify important new genes; 2) differential cloning of DNA regions present in one member of the pair but not in the other through genomic subtraction; 3) epidemiologic screening of the resulting DNA regions for association with UTI transmission, virulence, and maintenance using the investigators' well characterized collections of UTI and fecal E. coli isolates; and 4) characterization of potential UTI genes in the associated DNA regions. The investigators state that the discovery of additional factors involved in E. coli mediated UTI will facilitate the development of new strategies for the prevention and cure of UTI.